Published By : 13 Oct 2017 | Published By : QYRESEARCH
Amazon Web Services and Microsoft have collaborated to develop Gluon, a new deep learning interface that will make it easier for developers to create prototypes for machine learning models. The Gluon interface uses a Python API and a number of neural network components that come inbuilt with the system. The networks are optimized for use in modern applications of deep learning, which will make it easy for developers to create novel neural networks using relatively simple and, most importantly, concise code. The interface works with Apache MXNet at present, and is expected to receive an update in the near future that will enable support for Microsoft Cognitive Toolkit.
The main problem with developing new machine learning systems in recent years has been the high amount of work required on multiple fronts, which makes it unviable for many players in the tech sector. The specialized expertise required to make the most of deep learning technology is also lacking in several parts of the tech sector. According to Swami Sivasubramanian, VP of Amazon’s AI department, Gluon makes it easier to build machine learning systems to the extent that developers can use the platform to create deep learning systems just like mobile apps.
Microsoft’s Corporate VP of AI and Research echoed the sentiment, declaring Microsoft’s intentions to produce an environment where deep learning researchers can easily collaborate and turn ideas into reality. The high potential of machine learning to transform the way technologies operate has made Microsoft a key participant in the sector. Amazon, who recently announced a collaboration between their and Microsoft’s virtual assistants, Alexa and Cortana, are also understandably eager to make it big in the machine learning space.